Re: Mule bugs: misidentification (Latin-1 vs. Chinese), revert issues

Ar an deichiú lá de mí Deireadh Fómhair, scríobh Michael Sperber: OS X (which your headers tell me you’re on) uses something close to UTF-8 in normal form D--see http://developer.apple.com/qa/qa2001/qa1173.html for details--for file name encoding, where the canonical form for a precomposed character is the base character followed by a combining character. If you do C-x = on the box, it should tell you that it represents U+0308 COMBINING DIAERESIS; that it is not shown as a combined diaeresis is redisplay problem with XEmacs. The most common encoding of "ä" in UTF-8 is the representation of U+00E4, "\xc3\xa4". XEmacs handles that fine, which is why you won’t see this problem on Linux, where as a rule normalisation is not done. U+0061 followed by U+0308 (what you’re seeing: "a\xcc\x88" on disk) is preserved on reading and writing, which is good, and the OS X libraries automatically normalise when presented with non-normalised UTF-8 text, which is also good. So this is not a bug in the interaction of the locale with file names, rather in our handling of combining characters. Actual support for normalisation in the Unicode coding systems would also be nice to have, but is not amazingly relevant for this particular problem, since the OS does it already. -- On the quay of the little Black Sea port, where the rescued pair came once more into contact with civilization, Dobrinton was bitten by a dog which was assumed to be mad, though it may only have been indiscriminating. (Saki) _______________________________________________ XEmacs-Beta mailing list XEmacs-Beta(a)xemacs.org http://calypso.tux.org/cgi-bin/mailman/listinfo/xemacs-beta